Background/Question/Methods Multi-factor experiments are necessary for determining the potential of interactions among various global change factors in their effects on ecosystem functioning. The Old-Field Community Climate and Atmosphere Manipulation (OCCAM) experiment was initiated to assess single factor and interactive effects of [CO₂], warming, and altered soil moisture on old-field ecosystems. The factorial experiment consists of constructed plant communities (including C3 and C4 grasses, forbs, and legumes) in open-top chambers. After four years, we have found that the relative importance of single factor vs. interactive effects varied among years and depended on the response measured. Importantly, plant species differed in their responses to the climate change factors, resulting in changes in plant community composition. These shifts in plant dominance patterns probably have much larger effects on ecosystem functioning than direct effects of climate change factors on ecosystem processes. Effects on the soil system, which play a key role in long-term ecosystem responses and subsequent feedbacks to climate changes, may strongly depend on plant community composition. We tested the hypothesis that climate change-induced alteration of plant community composition prevails over direct climate change effects in affecting soil ecosystem functioning, as indicated by extracellular enzyme activities. In each chamber, we measured activities of nine enzymes that take part in degrading a variety of compounds involved in carbon and nutrient cycling. Samples were collected from under the two dominant plant species, Lespedeza cuneata (an exotic legume) and Festuca pratense (a native C3 grass), whose cover and biomass were significantly affected by the climate change treatments and thereby reflect the changes in plant community composition. Activities of enzymes involved in degradation of orthophosphate, sulfur, and lignin were lower in dry than in wet treatments.

Results/Conclusions Activities of enzymes involved in degradation of orthophosphate, sulfur, chitin and cellulose were lower under Lespedeza than under Festuca. [CO₂] and temperature had less effect. Multivariate analysis (to test treatment effects on enzyme ‘community’ activity) showed significant plant species x treatment interactions, which indicate that direct effects of climate change on the soil system can depend on, or be overruled by, changes in plant community composition. We propose that to better understand and predict the integrated ecosystem response to climate change, we should put more emphasis on climate change-induced shifts in plant community composition.